| |
Early Changes in Hepatitis C Virus (HCV) Levels in Response to Peginterferon and Ribavirin Treatment in Patients with Chronic HCV Genotype 1 Infection
|
| |
| |
"In this study as well as in previous reports, changes in HCV RNA levels at week 12 (referred to an early virological response) were more reliable in predicting ultimate response (or nonresponse), compared with changes at earlier time points"
"In the Virahep-C study, African American patients were less likely than white patients to become HCV RNA negative and less likely to achieve a sustained virological response [12]. The underlying cause of virological nonresponse and the reasons why it is more common among African American patients than among white patients were not clear. The current analyses demonstrated that these differences are fundamentally biologic and become apparent 24-48 h after starting therapy. A reasonable hypothesis to explain nonresponse to interferon-based antiviral therapy is the presence of a block or defect in the afferent or efferent limbs of intracellular interferon responses [19, 20]. In other studies of the Virahep-C cohort, poor responders were found to have a blunted expression of interferon-stimulated genes in peripheral blood mononuclear cells, both in the total number of genes induced and in the strength of the induction [25]. The basis for the blunted response is unknown but may be related to inhibition of interferon signaling pathways by HCV polypeptides, such as the NS3 protease [26] and the HCV core gene [27], or by a preexisting up-regulation of interferon-stimulated gene responses in nonresponding patients [28, 29]. The current study demonstrates that the block is found more commonly in African American patients, compared with white patients. Furthermore, this relatively poor interferon response is not limited to the liver: in this study, patients with a poor virological response also had a smaller decrease in platelet and white blood cell counts, although the degrees of difference in these changes were far less than for the HCV RNA level (or alanine aminotransferase and aspartate aminotransferase levels). Analysis of the virological, immunological, and host genetic factors that might account for poor antiviral responses to interferon therapy is an important focus for future research into attempts to improve response rates to current therapy for hepatitis C."
The Journal of Infectious Diseases April 15 2009;199:000-000
Jay H. Hoofnagle,1 Abdus S. Wahed,3 Robert S. Brown, Jr.,5 Charles D. Howell,2 and Steven H. Belle,3,4 for the Virahep-C Study Groupa
1Liver Disease Research Branch, Division of Digestive Diseases and Nutrition, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, and 2Division of Hepatology, University of Maryland School of Medicine, Baltimore, Maryland; Departments of 3Biostatistics and 4Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania; and 5Division of Gastroenterology and Hepatology, New York-Presbyterian Medical Center, New York, New York
Early changes in hepatitis C virus (HCV) RNA levels were assessed in a large cohort of African American and white patients with chronic hepatitis C due to HCV genotype 1 who underwent therapy with peginterferon alfa-2a and ribavirin in the Study of Viral Resistance to Antiviral Therapy of Hepatitis C (Virahep-C). Analyses were restricted to 341 patients who completed the first 28 days of therapy without dose modification.. HCV RNA levels decreased in virtually all patients, but the amount of the change varied markedly. The overall 28-day decrease in HCV RNA levels was at least as predictive of a sustained virological response as the first- or second-phase viral kinetics responses. Factors associated with a smaller decrease in the HCV RNA level between baseline and day 28 included African American race, higher initial HCV RNA level, more severe hepatic fibrosis, and higher body weight. African American patients with similar 28-day decreases in viral levels as white patients were still less likely to achieve a sustained virological response. These results suggest that racial differences in the response to antiviral therapy are due to greater unresponsiveness to intracellular actions of interferon in African American individuals and that standard doses of peginterferon and ribavirin may be suboptimal for patients with higher body weights.
Chronic hepatitis C affects 3.2 million Americans and is the major cause of cirrhosis, end-stage liver disease, and hepatocellular carcinoma in the United States [1]. The current recommended therapy for chronic hepatitis C is a combination of peginterferon and ribavirin given for 24 or 48 weeks, depending on the viral genotype [2-5].. For patients infected with hepatitis C virus (HCV) genotypes 2 or 3, combination therapy is highly effective, and a 24-week course of peginterferon and a reduced dose of ribavirin results in a sustained virological response in 75%-80% of patients [6, 7]. For patients with genotype 1, a full 48-week course of treatment with full doses of peginterferon and ribavirin leads to an undetectable HCV RNA level in only 65%-70% of patients and a sustained virological response in 45%-55% [6, 8-10]. Clearly, major challenges in hepatitis C therapy are the lack of a complete response to combination antiviral therapy, which occurs in approximately one-third of patients infected with HCV genotype 1, and the continued occurrence of relapse, regardless of the genotype of the infecting strain.
Both viral and host factors appear to be important in the virological response to combination therapy. In addition to viral genotype, higher baseline levels of HCV RNA are associated with lower rates of sustained virological response [6, 8-10]. Host factors that have been associated with lower sustained virological response include African American race, male sex, higher body weight, older age, diabetes and insulin resistance, more advanced degrees of hepatic fibrosis, and the amount of peginterferon and ribavirin received [2-12].
An important predictor of the ultimate response to interferon-based therapy has been early viral kinetics, shown by measurements of HCV RNA levels during the first few weeks of treatment [13-18]. Typically, HCV kinetics follow a biphasic pattern [13]. The first phase is characterized by a marked decrease in serum the HCV RNA level 24-48 h after the initial injection of interferon. The second phase is marked by a more gradual decrease in the HCV RNA level that continues over the ensuing several weeks and thereafter. The first-phase decrease represents the efficacy (ε) of interferon against HCV. The second-phase decrease reflects the rate of clearance (γ) of HCV-infected hepatocytes. Analysis of the first- and second-phase decreases in HCV RNA level may help to elucidate the nature and cause of nonresponse to interferon-based therapy for hepatitis C [19, 20].
This article includes analyses of the changes in HCV RNA levels during the first 28 days of therapy with standard doses of peginterferon and ribavirin in African American and white patients with chronic hepatitis C due to HCV genotype 1 who were participating in a prospective trial of combination therapy [12].
Results
Patient characteristics.A total of 401 patients were enrolled and treated with combination therapy in the Virahep-C study [12]. For analyses of the early viral kinetics, the 341 patients who had HCV RNA levels assessed at the designated time points and were treated with the full dose of peginterferon for the first 28 days were studied.
The baseline characteristics of the 341 patients, stratified by race, are shown in table 1. The average age was 51 years, and 66% were men; 154 were African American, and 187 were white. As previously reported for the entire cohort, African American patients had, on average, higher body weights and body mass indexes and were more likely to have a history of diabetes, insulin resistance, and hypertension, compared with white patients [12]. In addition, African American patients had lower initial alanine aminotransferase and aspartate aminotransferase values, despite similar histological scores for necroinflammatory activity and fibrosis. HCV RNA levels were similar in the 2 racial groups, but African American patients were more likely to have genotype 1b infection and less likely to have genotype 1a infection, compared with white patients.
Early changes in HCV RNA levels.Levels of HCV RNA decreased during the first 28 days of treatment in all except 2 patients (both of whom were African American), with a median decrease of 1.84 log10 IU/mL. The degree and pattern of decrease differed between African American patients and white patients (figure 1). Although baseline levels were similar, the 28-day decrease in the HCV RNA level was greater among white patients, compared with African American patients, and the racial difference in levels was already statistically significant by day 2 of treatment. By day 28, 22% of white patients but only 12% of African American patients were HCV RNA negative (defined as an HCV RNA level of <50 IU/mL; p=.02).
Assessment of early viral kinetics.
Changes in HCV RNA levels during the first 28 days of treatment were assessed in 3 ways: the first-phase kinetics response (days 0-2), the second-phase response (days 7-28), and the total decrease from baseline to day 28. Formal computation of first- and second-phase responses, using published formulas for efficacy (ε) and clearance rate (γ) [13], could not be done for most patients, largely because of a rebound in the HCV RNA level between the nadir on day 1 or 2 and the level on day 7, as well as inadequate sampling for fitting biphasic models. For these reasons, first-phase responses were calculated as the decrease in the log10 IU/mL HCV RNA level between baseline and day 1 or 2, whichever was greater. Second-phase responses were calculated as the maximum weekly decrease in the log10 HCV RNA level between days 7 and 28 or between days 14 and 28, whichever was greater. Second-phase responses could not be computed for 14 patients whose HCV RNA levels were not detectable by day 7 and were therefore not included in analyses of second-phase responses.
The median first-phase response, second-phase response, and overall log10 decrease in the HCV RNA level during the first 28 days of therapy are shown in table 2, stratified by selected categorical baseline factors. The median first-phase decrease was 0.75 log10 IU/mL among all patients and was greater among white patients (0.89 log10 IU/mL), compared with African American patients (0.60 log10 IU/mL; p<.01). Similarly, the median second-phase decrease was 0.46 log10 IU/mL/week overall and was greater among white patients (0.55 log10 IU/mL/week), compared with African American patients (0.38 log10 IU/mL/week; p<.01). Finally, the median decrease in HCV RNA level for the entire study sample over the first 28 days was 1.84 log10 IU/mL and was greater among white patients (2.56 log10 IU/mL), compared with African American patients (1.31 log10 IU/mL; p<.01). Other factors that were significantly associated with an improved kinetics response in one or both study groups included younger age, lower body weight, lower HOMA index, lower baseline HCV RNA levels, lower pretreatment score for fibrosis, and lack of steatosis, but not sex or viral genotype. Interestingly, baseline HCV RNA levels had a significant, negative association with the first-phase decrease (p=.001) but not with the second-phase decrease or the overall 28-day decrease. Body weight, HOMA index, and fibrosis scores were significantly associated with the overall 28-day decrease in the HCV RNA level, but the associations of these factors with the other viral kinetics measures were inconsistent: fibrosis scores were significantly associated only with first-phase responses and body weight only with second-phase responses. In multivariate analysis, factors independently associated with a 28-day decrease in the HCV RNA level were white race and clinical features associated with less severe degrees of hepatic fibrosis (i.e., higher white blood cell and platelet counts), but not sex, body weight, or HOMA index (data not shown).
Early changes in HCV RNA levels and sustained virological response.
The viability of using first- or second-phase responses and overall 28-day decrease in HCV RNA levels for predicting achievement of a sustain virological response was analyzed by computing receiver operating characteristic (ROC) curves (figure 2), with adjustment for other predictive factors [12]. All 3 measurements of early viral kinetics predicted achievement of sustained virological response well (area under the curve, >0.80 for all), but the area under the curve was significantly greater for the second-phase decrease (0.876) and the 28-day decrease (0.879), compared with the first-phase kinetics (0.806; p<.001for both comparisons). The difference in areas under the curve between the second phase and 28-day decreases was not statistically significant (p=.61). In further analyses, the 28-day decrease was used.
Early viral kinetics and sustained virological response rates, by race and sex.
Sustained virological response rates, stratified by the 28-day decrease in the log10 HCV RNA level, are shown for men, women, African American patients, and white patients in table 3. Patients with a decrease of <1 log10 IU/mL at 28 days had only a 13% chance of achieving a sustained virological response. With increasing 28-day decreases in HCV RNA levels, the likelihood of achieving a sustained virological response increased, such that 85% patients with a decrease of >4 log10 IU/mL had a sustained virological response. The sustained virological response rate for 60 patients who tested negative for HCV RNA at week 4 (frequently referred to as rapid virological responders [4, 7]) was 77%, a relatively low response rate that perhaps reflects the high proportion of African American patients and genotype 1 infections in this study. After levels of 28-day log10 decreases were controlled for, sustained virological response rates were significantly less among African American patients, compared with white patients (p=.037). After adjustment for 28-day decreases, sustained virological response rates were higher among women, compared with men (p=.023), particularly at lower levels of decrease. Thus, among subjects with a 28-day decrease of <2 log10 IU, 16% of men and 36% of women achieved a sustained virological response (p=.01), whereas among those with a decrease of 2 log10 IU, the difference in sustained virological response rates was less (65% in men vs. 72% in women; p=.39).
After other factors shown to be associated with a sustained virological response were controlled for, higher sustained virological response rates were found among women, compared with men, and among white subjects, compared with African American subjects, over the range of log10 IU/mL 28-day decreases in HCV RNA levels (figure 3). These results demonstrate that initial viral kinetics did not completely account for differences in response rates between women and men and between white patients and African American patients.
On treatment factors and 28-day decreases in HCV RNA levels.
Treatment-related factors associated with 28-day decreases in HCV RNA levels were identified (table 4). Patients with greater 28-day decreases in HCV RNA levels were also more likely to have greater decreases in serum alanine aminotransferase and aspartate aminotransferase levels and, to a lesser extent, white blood cell and platelet counts. Worsening symptoms, as assessed by visual analogue scales, were not significantly associated with a change in the HCV RNA level during the 28-day period. Peginterferon levels at 28 days were higher among patients with an HCV RNA level that decreased by >2 logs, compared with patients with a smaller decrease, but the differences were slight. After body weight and peginterferon levels were controlled for, the associations between 28-day decreases in HCV RNA levels and decreases in white blood cell counts and platelet counts were still statistically significant (data not shown). Thus, patients with greater degrees of viral suppression at 28 days also had greater degrees of suppression in the bone marrow, as assessed by white blood cell and platelet counts.
Discussion
In this prospective study, the early changes in HCV RNA levels during receipt of antiviral therapy for chronic HCV genotype 1 infection were highly predictive of ultimate outcome. Thus, the likelihood of achieving a sustained virological response was far greater among patients with a decrease in the HCV RNA level of >2 log10 IU/mL by 28 days (68%) than among those in whom the decrease was less (22%). Furthermore, patients in whom the HCV RNA level decreased by >4 log10 IU/mL by 28 days had an 85% chance of achieving a sustained virological response, which is similar to percentages among patients infected with genotype 2 or 3 [2, 6, 7]. Importantly, however, the 28-day changes in HCV RNA levels (as well as the first- and second-phase viral kinetics) were not completely reliable in predicting achievement of a sustained virological response. Thus, 6 patients (15%) in whom the 29-day decrease was >4 log IU/mL did not have a sustained virological response, and, contrariwise, 44% of those in whom the 28-day HCV RNA level decreased by 2 log10 IU/mL nevertheless achieved a sustained virological response. In this study as well as in previous reports, changes in HCV RNA levels at week 12 (referred to an early virological response) were more reliable in predicting ultimate response (or nonresponse), compared with changes at earlier time points [2-5, 12].
The results of the current study also showed that the decreases in HCV RNA levels during peginterferon and ribavirin therapy did not fit the published kinetics models for first- and second-phase viral responses [13]. The findings for responders with the greatest 28-day decreases in the HCV RNA level were most compatible with findings from published models, whereas viral kinetics in poor responders often demonstrated changes that could not be characterized using standard differential equations. In most cases, these difficulties were caused by the rebound in the HCV RNA level that occurred between days 2 and 7 of therapy. In this respect, it should be stressed that the studies on viral kinetics that first derived formulas for first- and second-phase decreases in HCV RNA levels used daily injections of standard interferon alfa in fairly high doses [13]. Thrice weekly standard interferon therapy and once weekly peginterferon treatment yield different patterns of decreases in the HCV RNA level that may not match those of previously published models [15-18].
In this and other studies from the United States, overall viral kinetics responses were poorer among African American patients, compared with white patients [12, 16]. The clinical and laboratory features that differed between the 2 races (such as body weight, HOMA index, and presence of diabetes) could not fully explain these differences. Factors other than race that were associated with poor 28-day antiviral responses included greater initial virus level, liver fibrosis, and greater body weight. Importantly, however, initial viral kinetics did not fully account for racial differences in sustained virological response rates. Thus, for a given 28-day decrease in the HCV RNA level, white patients were still more likely than African American patients to have a sustained response.
In analyses of sustained virological response rates, women had a higher rate of response to therapy, compared with men. As previously reported, for the entire Virahep-C cohort, the sustained virological response rate was 13% higher among women, compared with men, in both racial groups (sustained virological response rates, 24% among African American, 35% among African American women, 47% among white men, and 61% among white women) [12]. However, in analyses of early viral kinetics, higher body weight rather than sex remained significantly associated with greater 28-day decreases in HCV RNA levels. The association of body weight rather than sex with initial changes in HCV RNA levels remained significant even after controlling for other factors.
One interpretation of the association between early viral response and body weight is suboptimal dosing of peginterferon and/or ribavirin. Thus, some overweight or obese patients who did not respond to or experienced relapse after receipt of peginterferon and ribavirin therapy may not have responded to therapy, because of suboptimal dosing. These results suggest that further studies on antiviral dosing that control for body weight are warranted.
In the Virahep-C study, African American patients were less likely than white patients to become HCV RNA negative and less likely to achieve a sustained virological response [12]. The underlying cause of virological nonresponse and the reasons why it is more common among African American patients than among white patients were not clear. The current analyses demonstrated that these differences are fundamentally biologic and become apparent 24-48 h after starting therapy. A reasonable hypothesis to explain nonresponse to interferon-based antiviral therapy is the presence of a block or defect in the afferent or efferent limbs of intracellular interferon responses [19, 20]. In other studies of the Virahep-C cohort, poor responders were found to have a blunted expression of interferon-stimulated genes in peripheral blood mononuclear cells, both in the total number of genes induced and in the strength of the induction [25]. The basis for the blunted response is unknown but may be related to inhibition of interferon signaling pathways by HCV polypeptides, such as the NS3 protease [26] and the HCV core gene [27], or by a preexisting up-regulation of interferon-stimulated gene responses in nonresponding patients [28, 29]. The current study demonstrates that the block is found more commonly in African American patients, compared with white patients. Furthermore, this relatively poor interferon response is not limited to the liver: in this study, patients with a poor virological response also had a smaller decrease in platelet and white blood cell counts, although the degrees of difference in these changes were far less than for the HCV RNA level (or alanine aminotransferase and aspartate aminotransferase levels). Analysis of the virological, immunological, and host genetic factors that might account for poor antiviral responses to interferon therapy is an important focus for future research into attempts to improve response rates to current therapy for hepatitis C.
Methods
The Study of Viral Resistance to Antiviral Therapy of Hepatitis C (Virahep-C) was a multicenter study of the combination of peginterferon and ribavirin for treating patients with chronic hepatitis C. The study was designed to assess the rates of response among African American and white patients and to help to elucidate the causes of nonresponse to interferon-based therapy. The aims, design, and results of this study have been described elsewhere [12]. Adult patients aged 18-70 years with chronic hepatitis C due to HCV genotype 1 who had not been previously treated were eligible for enrollment. Exclusion criteria included decompensated liver disease and other major complicating illnesses. All patients were required to have been born in the United States and to have self-identified as either "African American/black" or "Caucasian/white" and not "Both" or "Other." Approximately 50 patients, equally divided between African American individuals and white individuals, were enrolled at each of 8 clinical centers. All patients gave written informed consent, and the protocol and consent forms were approved by the institutional review boards of the participating centers.
All patients received peginterferon alfa-2a (Pegasys [Roche Pharmaceuticals]) in a dose of 180 ƒÊg weekly and ribavirin (Copegus [Roche]) in a dose of 1000 (for patients with a body weight of <75 kg) or 1200 mg (for those with a body weight of >75 kg) daily. The first doses of peginterferon and ribavirin were administered under observation. Patients were asked to return to provide blood samples and undergo clinical evaluations after 24 and 48 h (the first one-third of the enrolled patients at each location also returned after 72 h); 7, 14, and 28 days later; and every 4 weeks thereafter. Importantly, because the samples from days 7, 14, and 28 were taken before the weekly injections, levels detected in these samples represented trough values. Therapy was continued for at least 24 weeks. Patients who remained HCV RNA positive at 24 weeks were considered nonresponders, and their therapy was discontinued. Patients who were not HCV RNA positive at 24 weeks continued to receive therapy for an additional 24 weeks. Patients were followed for at least 24 weeks after stopping therapy.
Standard clinical and anthropometric measurements were obtained at baseline.. Symptoms were assessed by a visual analog scale that provided assessment of the degree of fatigue and sense of well-being on a scale of 1-100. Insulin sensitivity was assessed by the homeostasis model assessment (HOMA) index, calculated as follows: [(fasting insulin level in ƒÊU/mL) (fasting glucose level in mg/dL/18)/22.5] [21]. HCV RNA testing was done at a central laboratory (SeraCare BioServices). HCV RNA levels were measured by the Cobas Amplicor Hepatitis C Virus Monitor Test, v2.0 (Roche Molecular Diagnostics), which has a lower limit of detection of 600 IU/mL. Samples that tested negative for HCV by the quantitative assay at weeks 24 and 48 were retested in duplicate, using the Amplicor assay, which has a lower limit of detection of 50 IU/mL (Roche). Samples that were nonreactive by quantitative assays but reactive by qualitative assays were assigned an HCV RNA level of 325 IU/mL (which is midway between the lower level of detection for qualitative and quantitative assays), whereas samples testing negative for HCV by both assays were assigned an HCV RNA level of 25 IU/mL (which is midway between zero and the lower level of detection for the qualitative assay). Samples for which only quantitative results were known to be below the lower level of detection were assigned a value of 300 IU/mL (midway between zero and the lower level of detection for the quantitative assay). HCV genotyping was performed using a line probe hybridization assay (Versant HCV genotype assay [Bayer]). Serum peginterferon concentrations were measured by enzyme-linked immunoassay (Roche Molecular Diagnostics) [22]. Liver biopsy specimens were evaluated by a single study pathologist and scored under code for necroinflammatory activity and fibrosis, using a modification of the Ishak histological activity and fibrosis scoring systems [12, 23].
Baseline characteristics across the 2 racial groups were compared using the ƒÔ2 test for association, with continuity correction for differences in proportions. The nonparametric Wilcoxon rank sum test or the t test (as appropriate) was used to compare continuous variables. Differences in 28-day log10 decrease in the HCV RNA level across categorical baseline characteristics were tested for statistical significance by the Kruskal-Wallis test. The associations between the 28-day log10 decrease and other baseline characteristics and changes in biochemical and hematologic factors as well as peginterferon levels and symptoms were assessed by univariable and multivariable linear regression analyses. The associations between the first-phase response, the second-phase response, and the 28-day log10 decrease and achievement of a sustained virological response were assessed through logistic regression models after adjustment for potential confounders [24].. The probability of achieving a sustained virological response on the basis of these early changes in the HCV RNA level was compared using receiver operating characteristic (ROC) curves. Differences in the proportions of participants with a sustained virological response, stratified by 28-day log10 decrease in the HCV RNA level, were compared using a test for trend (Jonckheere-Terpstra test). After levels of 28-day log10 decrease were controlled for, the differences in sustained virological response rates between men and women and between African American and white subjects were compared using the Cochran-Mantel-Haenszel test. Statistical analyses were performed using SAS, version 9.1.3 (SAS Institute).
|
|
| |
| |
|
|
|
